Introduction to Compositional Data in Archaeology
- Understanding Compositional Data
- Definition and examples of compositional data in archaeology (e.g., proportions of different materials, chemical compositions).
- Example: Typical archaeological datasets that include compositional data.
- Challenges in Analysing Compositional Data
- Issues with relative proportions and the “closed” nature of compositional data.
- Example: Limitations of traditional statistical methods on compositional data.
Basic Concepts in Compositional Data Analysis
- Overview of methods in Multivariate statistics
- Log-Ratio Transformations
- Introduction to log-ratio transformations (CLR, ILR, ALR) for compositional data.
- Example: Applying a centred log-ratio (CLR) transformation in R (e.g.
nexus::transform_clr).
- Visualizing Compositional Data
- Techniques for visualizing compositional data (ternary plots, bar charts).
- Example: Creating a ternary plot using the
isopleuros::ternary_plot and ggtern.
Exploratory Data Analysis
- Principal Component Analysis (PCA)
- Introduction to PCA tailored for compositional data.
- Example: Performing PCA on compositional data using
tesselle::nexus::pca.
- Biplots and screeplots
- Example: Plotting PCA results as a biplot and add visualisation aids using
tesselle::dimensio functions.
- Cluster Analysis
- Overview of clustering techniques for exploring groupings in compositional data.
- Example: Applying hierarchical clustering on transformed compositional data using
tesselle.
ggplot2: dendrograms with ggraph
- Correspondence Analysis
- Correspondence Analysis for compositional data.
- Example: Performing Correspondence Analysis using
tesselle::ca.
- Case Study: Analysis of Compositional Data from Archaeological Sites
- Step-by-step walkthrough of an exploratory analysis of compositional data.
- Example: Analysing chemical compositions of ceramics or soils using
tesselle.
- Q&A and Troubleshooting
- Addressing challenges in visualizing and analyzing compositional data in archaeological research.